The effect of raindrop clustering on collision‐induced break‐up of raindrops

The impact of raindrop clustering on raindrop mean free paths, mean times between raindrop collisions, and raindrop collision rates is examined using prior observations of raindrop pair cross‐correlation functions and raindrop size distributions (RSDs). Variations in observed RSDs cause collision rates to vary by more than a factor of 2.5 for identical rain rates; effects due to raindrop clustering also increase collision rates by 2.5 times or more. High uncertainty in the factor of 2.5 and its variability exists because observations of raindrop inter arrival times, upon which observations of clustering are based, are relatively rare. Clustering increases collision rates because large raindrops in preferred or clustered regions overtake smaller drops in the same regions. The form of the RSD has a significant impact on the collision frequency because the spread in raindrop fall speeds, and hence the ability of large drops to overtake small ones, depends highly on the RSD. Depending on the form of the RSD that develops at precipitation onset, numerical models suggest that there can be sufficient raindrop interactions to approach equilibrium distributions generated from the opposing forces of coalescence and break‐up for rain rates higher than about 50 mm h −1 . However, equilibrium distributions do not occur in nature, most likely because the large spread in fragment size distributions generated for a specific pair of colliding raindrops causes substantial local variations in RSDs. Significant sampling times under conditions of heavy rain would be required to return to the form of an equilibrium distribution from statistical effects alone, if it indeed existed. Copyright © 2004 Royal Meteorological Society